Shreya Tripathi is a second-year B.A., LL.B. (Hons.) student at IILM University, Greater Noida. Her academic interests centre on constitutional law, with a growing inclination toward issues at the intersection of rights, governance, and contemporary legal challenges. She has gained practical exposure to the legal system through her internship experience with a law firm, where she engaged with varied aspects of legal research and drafting. Additionally, her visits to courts have provided her with valuable insight into courtroom procedure and the functioning of the judicial process, further strengthening her interest in litigation and public law.
Every time a court in India issues a judicial order against a deepfake or the Ministry of Electronics and Information Technology (MeitY) announces changes to existing laws, the same phrase pops up in the headlines: India is “cracking down” on deepfakes. However, the more laws are introduced, the clearer it becomes that the progress is not really moving forward it’s just adding more layers to old rules, not actually advancing legal principles. In this article, I will argue that simply having a law that applies to a deepfake in India isn’t enough. Such a law is always present, either in the Information Technology Act of 2000, the Bharatiya Nyaya Sanhita of 2023 (BNS), or the amended Intermediary Guidelines[1]. What is needed is that three things happen at the same time: identifying who created the deepfake, stopping any harm before it becomes too serious, and making sure this is done equally well for someone working in a textile factory in Ranchi and a movie star in Mumbai.
THE PATCHWORK THAT EXISTED BEFORE ANYONE SAID THE WORD “DEEPFAKE”
Prior to the publication of the draft amendments in October 2025. However, there were some broader rules that were being used to address the issue. Section 66C of the IT Act makes it a crime to steal someone’s identity[2]. Section 66D deals with cheating by pretending to be someone else using a computer[3]. Section 66E covers the creation[4], sharing, or sending of images of a person’s private parts without their permission. Sections 67 and 67A are about publishing content that is obscene or sexually explicit. The BNS has a law that stops forgery, cheating, defamation, and making false statements that cause public disturbance[5]. The Digital Personal Data Protection Act, 2023 strengthens protection against non-consensual deepfakes by safeguarding biometric and other personal data through strict penalties for misuse.
It was not the lack of pertinent text, rather, it was the fact that nowhere in any of these provisions did it talk about “synthetic” anywhere, nowhere did it consider the possibility of an unreal video looking real, and nowhere was there a provision to take down the content prior to it becoming viral. In other words, the prosecution was trying to force a new problem through old law. This was exactly what happened in the widely reported Andhra Pradesh deepfake incident, where a young engineer allegedly altered a woman’s image using a generative adversarial network and was proceeded against under provisions relating to forgery and impersonation, illustrating how existing criminal law was being adapted to address AI-generated harms[6]
COURTS FILLED THE VACUUM FIRST
In the absence of any legislation on this issue from Parliament or government, the Delhi High Court became the first judicial forum to tackle the problem of deepfake damages. It relied upon both common law principles as well as constitutional provisions concerning the right of the person to control how his/her name, voice, image, and likeness is used, especially if he/she is a respectable personality in the commercial world. Actors and celebrities who knew that the AI technology was reproducing them in advertisements of certain products, making comments on politics, or being featured in fake videos approached the court for injunctions against the damage done. The Delhi High Court granted such relief in cases including Amitabh Bachchan v Rajat Nagi[7] and Anil Kapoor v Simply Life India[8], recognizing and protecting personality and publicity rights against AI-enabled misuse.
It is important to note such a quick response in a legal manner because it provided assistance where there was no legal mechanism. But this quick legal response also highlights how inefficient the legal mechanism was through which this has been accomplished. Litigation through rights-based cases can be facilitated only when individuals are celebrities and have the capacity to bring the case forward in an urgent manner. It makes sense that when a person is considered to be in the category of a celebrity, then he/she may consider that his/her public image along with his/her voice has some economic worth and needs to be protected. But a common individual does not have any worth at all.
THE REGULATORY TURN: OCTOBER 2025 TO FEBRUARY 2026
The legislative act was a stepwise procedure. In 2025, on 22 October, MeitY published draft amendments[9] to the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021. These amendments included the definition of synthetically generated information and the obligation of platform operators to obtain confirmation from the users, verification of the content using adequate technologies, and labelling all such contents. In the case of visual content, the label has to be placed no less than one tenth of the screen; meanwhile, regarding audio content, it is required to be done at the beginning of the video. The final amendments were notified on 10 February 2026 and came into force on 20 February 2026. Most importantly, the amendments supplemented the existing notice-and-takedown framework by imposing proactive due diligence obligations on intermediaries dealing with synthetically generated information. Instead, the social media operators have to undertake certain steps in order to retain immunity granted to them under the provisions of Section 79 of the IT Act[10].
On 10 February 2026, MeitY notified another amendment, which came into force on 20 February 2026, and this time, it worked towards improving the system further by making clear what kinds of edits, like noise reduction, formatting, and compression, were considered normal and honest, thereby ensuring that everyday content production does not come under regulation because of the misunderstanding of what a manipulation is. The amendment also added more stringent deadlines and provided for multiple authorized officers to be able to give the instruction for removal of content, as harmful synthetic content spreads at a rapid pace.
On the other hand, in 2024, there was another private member’s bill, the Regulation of Deepfake Bill, 2024, presented in the Lok Sabha[11]
This bill proposed punishment for the creation of deepfakes without consent, consent and watermarking for deepfakes, and creation of a Deepfake Task Force. Private members’ bills hardly become laws in India, very few of them have become laws since India’s independence, but the very fact that this bill was introduced shows the seriousness of the matter.
When combined, all of these actions comprise the most comprehensive regulatory framework for addressing a particular form of online harm that India has seen in some time. When taken alone, it is a well-reasoned approach to addressing an actual problem, one where the platforms for the first time will have clear legal obligations instead of being subject to ad hoc rulings from court cases.
WHERE THE GAPS STILL SIT
There are several main issues that arise based on the above-stated three-part test from the very beginning of this blog. Even after that, attribution remains a problem because both detection technology and labelling have been found to be unreliable until now and labels can even be deleted.
“Synthetic Information” has a very wide definition and includes parody and satire. The powers granted to the executive authorities to determine what qualifies as unlawful content creates problems under the Constitution in light of the case of Shreya Singhal v Union of India.
Finally, there are no remedies for the general public, particularly women who become victims of sexually explicit deepfakes.
WHAT “SUFFICIENT” WOULD ACTUALLY REQUIRE
The statement that it has not undertaken any substantial action or the actions that were undertaken between October 2025 to February 2026 are cosmetic in nature does not imply that India has done nothing but that there being criminal provisions or laws or labelling laws is not necessarily an indication of adequacy. A sufficiently strong framework should comprise at least the choice of seeking justice within a specified period and also the imposition of fines that are fixed irrespective of the fact that the person who complains is a victim or not. Additionally, there should be proper definitions as far as the identification of harmful AI-generated contents and their uses in artistic and journalism context is concerned and such definitions need to be independently reviewed by some third-party organization that can also remove the said content.
In two years, India has achieved a lot, the country started out with no specific regulations regarding the use of deepfake material, and now India boasts one of the best regulatory frameworks regarding deepfakes in the whole world. The framework is better than those of such powerful states as the European Union and the United States (with regard to their proposals), since it has stricter demands with regard to labelling and taking down the material than any other country. However, the mere fact of attempts made cannot solve all problems. And the answer to the question posed in the article is still “not yet”.
[1] Information Technology Act 2000; Bharatiya Nyaya Sanhita 2023; Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules 2021.
[2] Information Technology Act 2000, s 66C.
[3] Information Technology Act 2000, s 66D.
[4] Information Technology Act 2000, s 66E.
[5] Bharatiya Nyaya Sanhita 2023, ss 336–340 (forgery), ss 318–319 (cheating), ss 356–357 (defamation), as applicable
[6] The Indian Express, ‘Andhra engineer arrested for creating AI-generated obscene images’
[7] Amitabh Bachchan v Rajat Nagi 2022 SCC OnLine Del 4110.
[8] Anil Kapoor v Simply Life India 2023 SCC OnLine Del 6914.
[9] Ministry of Electronics and Information Technology, Draft Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules (22 October 2025).
[10] Information Technology Act 2000, s 79.
[11] The Regulation of Deepfake Bill, 2024 (Private Member’s Bill)
